&get_box_var;

ORIGINAL ARTICLE

Microbiota Promotes Chronic Pulmonary Inflammation by Enhancing

IL-17A and Autoantibodies
Koshika Yadava1,2, Céline Pattaroni1, Anke K. Sichelstiel1, Aurélien Trompette1, Eva S. Gollwitzer1, Olawale Salami1,
Christophe von Garnier3,4, Laurent P. Nicod1, and Benjamin J. Marsland1
1
Service de Pneumologie, Faculté de Biologie et de Médecine, Centre Hospitalier Universitaire Vaudoise–Université Lausanne
Lausanne, Switzerland; 2Division of Infectious Diseases, Department of Medicine, Stanford University School of Medicine, Stanford,
California; 3Department of Respiratory Medicine, Inselspital, Bern University Hospital, Bern, Switzerland; and 4Department of Clinical
Research, University of Bern, Bern, Switzerland

Abstract richness and diversity were decreased in LPS/elastase–treated

Rationale: Changes in the pulmonary microbiota are associated
with progressive respiratory diseases including chronic obstructive
pulmonary disease (COPD). Whether there is a causal relationship
between these changes and disease progression remains unknown.
Objectives: To investigate the link between an altered microbiota
and disease, we used a murine model of chronic lung inflammation
that is characterized by key pathological features found in
COPD and compared responses in specific pathogen–free (SPF)
mice and mice depleted of microbiota by antibiotic treatment or
devoid of a microbiota (axenic).
Methods: Mice were challenged with LPS/elastase intranasally over
4 weeks, resulting in a chronically inflamed and damaged lung.
The ensuing cellular infiltration, histological damage, and decline
in lung function were quantified.
Measurements and Main Results: Similar to human
disease, the composition of the pulmonary microbiota was
altered in diseased animals. We found that the microbiota
mice, with an increased representation of the genera
Pseudomonas and Lactobacillus and a reduction in Prevotella.
Moreover, the microbiota was implicated in disease development
as mice depleted, or devoid, of microbiota exhibited an
improvement in lung function, reduced inflammation, and
lymphoid neogenesis. The absence of microbial cues markedly
decreased the production of IL-17A, whereas intranasal
transfer of fluid enriched with the pulmonary microbiota
isolated from diseased mice enhanced IL-17A production
in the lungs of antibiotic-treated or axenic recipients.
Finally, in mice harboring a microbiota, neutralizing
IL-17A dampened inflammation and restored lung
function.
Conclusions: Collectively, our data indicate that host–microbial
cross-talk promotes inflammation and could underlie the chronicity
of inflammatory lung diseases.
Keywords: microbiome; COPD; autoimmunity; IL-17

The rise in the prevalence of chronic
obstructive pulmonary disease (COPD)
is a global health concern (1). Exposure
to cigarette smoke is the most widely
associated environmental risk factor for the
development of the disease (2), which
presents as chronic bronchitis and
emphysema that lead to progressive and
irreversible airflow limitation (3, 4).
Although patients can manifest mild to
severe disease, as defined by the degree of
airflow obstruction, the signals leading to
increased severity and progression remain
unclear (3, 5). Despite the heterogeneity in
human disease, it is apparent that aberrant
inflammatory responses substantially
contribute to the decline in lung function
(5–7).

( Received in original form April 19, 2015; accepted in final form December 1, 2015 )
Funded by the Swiss National Science Foundation grant CRSII3_141875 (B.J.M. and C.v.G). K.Y. is currently supported by the Swiss National Science
Foundation early postdoc mobility grant.
Author Contributions: Conceived and designed the experiments: K.Y. and B.J.M. Performed the experiments and analyzed data: K.Y., A.K.S., A.T., O.S.,
E.S.G., C.P., and B.J.M. Prepared the manuscript: K.Y., B.J.M., L.P.N., and C.v.G.
Correspondence and requests for reprints should be addressed to Koshika Yadava, Ph.D., Division of Infectious Diseases, Department of Medicine, Beckman
Building B237, 279 Campus Drive, Stanford University School of Medicine, Stanford, CA 94305. E-mail: koshika@stanford.edu
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org
Am J Respir Crit Care Med Vol 193, Iss 9, pp 975–987, May 1, 2016
Copyright © 2016 by the American Thoracic Society
Originally Published in Press as DOI: 10.1164/rccm.201504-0779OC on December 2, 2015
Internet address: www.atsjournals.org

Yadava, Pattaroni, Sichelstiel, et al.: Microbiota Sustains Chronic Pulmonary Inflammation 975

At a Glance Commentary
Scientific Knowledge on the
Subject: The airway microbiota is
different in healthy individuals and
patients with chronic obstructive
pulmonary disease. However, its
function is unknown.
What This Study Adds to the
Field: This study provides
mechanistic insight into the function of
the airway microbiota in a mouse
model of chronic obstructive
pulmonary disease.
Intriguingly, not all people exposed to
similar levels of cigarette smoke develop
the disease (3). Moreover, although
smoking cessation does lead to an
improvement in lung function in moderate
disease, in some cases it does not impact
disease progression (8). One could
speculate that although environmental
triggers enhance the disease, in susceptible
individuals the established disease can
progress through a self-sustained
inflammatory cycle. Given the limitations
of current therapies to arrest COPD
progression, it is important to identify the
factors that can facilitate the establishment
of the chronic inflammatory pathways and
favor the disease. One such factor could be
the microbiota. Previously considered sterile,
breakthroughs in non–culture-based
detection methods have provided evidence
that the respiratory tract harbors a
microbiota (9–18). Notably, the microbiota
in the respiratory tract of healthy volunteers,
smokers without COPD, and patients with
severe disease are distinct (9, 10, 12–14).
Although microbial dysbiosis has been
shown to increase susceptibility to chronic
diseases such as inflammatory bowel disease
(19, 20), it is currently unknown whether
such a phenomenon contributes to COPD.
Another emerging hypothesis suggests
there is an autoimmune component driving
the progression of COPD (5, 21, 22).
Although controversial, there is both
circumstantial and direct evidence
implicating autoimmune mechanisms in
COPD pathogenesis (5, 21, 22). Lymphoid
follicles play a pathological role in several
autoimmune diseases by mediating the
generation of self-reactive antibodies (23).
Although lymphoid follicles themselves
976 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 9 | May 1 2016
have been consistently observed in both
animal models and humans with COPD
(24–27), there are conflicting reports
regarding the production of self-reactive
antibodies (28–32); thus, the autoimmune
concept of COPD remains open and might
reflect a subtype of the disease. Incidentally,
cues from the microbiota have been
reported to support lymphoid neogenesis
(33) and the production of self-reactive
antibodies (34) and are implicated in the
pathogenesis of a variety of autoimmune
models (34–36). Whether there could be a
similar link between the pulmonary
microbiota and autoimmune mechanisms
in COPD is yet to be established.
We aimed to elucidate the role of the
microbiome in COPD using a murine model
(37), which mimics key features of human
disease. We found that akin to what has
been seen in humans, the airway microbiota
differs between healthy and diseased mouse
lungs. We report that the microbiota has
functional implications, as in its absence
disease development is abrogated.
Methods
Mice
BALB/c mice were purchased from Charles
River, l’Arbresele, France. All animal
experiments were performed according to
institutional guidelines and Swiss federal
and cantonal laws on animal protection.
Model of Chronic Pulmonary
Inflammation
Mice were exposed intranasally to a mixture
of 7 mg LPS from Escherichia coli O26:B6
(Sigma-Aldrich, St. Gallen, Switzerland)
and 1.2 U porcine pancreatic elastase
(Elastin Products Company, Owensville,
MO) in 100 ml once a week over 4 weeks.
Terminal readout was performed 1 week
after the last challenge.
Assessment of Pulmonary Lung
Function
Lung compliance and FEV/FVC parameters
were measured by Snapshot and NPFE
perturbations using the FlexiVent invasive
airway mechanics system from Scireq
(Montreal, Canada). Mice were anesthetized
by administering 100 mg/kg ketamine
(Ketasol-100; Graeub) intramuscularly and
50 mg/kg pentobarbital (Esconarkon;
Streuli Pharma, Uznach, Switzerland)
intraperitoneally. Subsequently, mice were
ORIGINAL ARTICLE
tracheotomized and mechanically
ventilated at a rate of 450 breaths/min and a
tidal volume of 10 ml/kg bodyweight.
Antibiotic Treatment
For depleting microbiota before the start
of the experiment, mice were given 10%
enrofloxacin (Baytril) in drinking water for
2 weeks followed by amoxicillin/clavulanic
acid (Co-Amoxi-Mepha) in the drinking
water for another 2 weeks. During the course
of the experiment mice were maintained on
Co-Amoxi-Mepha treatment.
16S rDNA preparation from
bronchoalveolar lavage fluid and
sequencing. Bronchoalveolar lavage (BAL)
was performed in 1 ml volume of sterile
phosphate-buffered saline (PBS) and
collected in a 2 ml Biopure tube (Eppendorf,
Basel, Switzerland). Negative control was
obtained by flushing 1 ml of sterile PBS into
the BAL tubing system and processed
identically to the mouse BAL fluid (BALF)
samples. BALF was directly centrifuged at
14,000 3 g for 4 minutes at 48 C and
processed in sterile conditions under a
laminar flow hood. Pellets were incubated
with 9,000 U of Ready-Lyse Lysozyme
(Epicentre, Hessisch Oldendorf, Germany)
for 1 hour at 378 C. Twenty microliters of
proteinase K in 200 ml AL buffer (QIAamp
DNA Mini Kit) was then added to the
solution and incubated for 30 minutes at
568 C with shaking. After this step, BALF
from four mice coming from two different
cages were pooled and further processed with
QIAmp DNA Mini Kit, according to the
manufacturer’s protocol. DNA was eluted
in 30 ml DNase/RNAse-free water (Sigma-
Aldrich). Polymerase chain reaction (PCR)
for the 16S rDNA library preparation was
performed using modified 27F and 338R
universal primers, which target the V1-V2
hypervariable region of the 16S rDNA gene.
Primers were as follows:
27F:
59-AATGATACGGCGACCACC
GAGATCTACACTATGGTAATTCC
AGMGTTYGATYMTGGCTCAG-39
and
338R:
59-CAAGCAGAAGACGGCA
TACGAGATNNNNNNNNNNNN
AGTCAGTCAGAAGCTGCCTCCCG
TAGGAGT-39,
where Illumina adaptor sequences
are bold, linkers are italicized, and the
NNNNNNNNNNNN sequence represents
ORIGINAL ARTICLE
the sample-specific molecular identifier tag
barcodes. The temperature cycles were set as
follows: 3 minutes of initial denaturation at
948 C, 40 cycles of 30 seconds of denaturation
at 948 C, 30 seconds of annealing at 568 C,
1.3 minutes of extension at 728 C, and a final
step of 5 minutes at 728 C. PCRs were
performed in triplicate, and one additional
PCR negative control without DNA
template was added for each reaction. Each
reaction consisted of 8 ml of DNA template,
0.44 ml of each 27F and 338R primers at
10 mM, 2 ml of AccuPrime buffer II, 0.09 ml
of AccuPrime Taq DNA polymerase high
fidelity (Invitrogen), and DNase/RNAse-
free water (Sigma-Aldrich) to reach a total
volume of 20 ml. After amplification, PCR
product triplicates were pooled before
visualization on a 1.5% agarose gel. The
bands of the specific size (approximately
300 bp) were cut and purified with QIAquick
gel extraction kit (Qiagen, Hombrechtikon,
Switzerland). Recovered DNA was then
quantified using the Quant-iT PicoGreen ds
assay (Life Technologies, ThermoFisher,
Waltham, MA), and amplicons were pooled
in a single tube in equimolar amounts.
Sequencing was performed on an Illumina
MiSeq platform using MiSeq reagent kit
V2-500 (pair-end, 2 3 250).
16S rDNA sequencing analysis.
Sequences obtained were processed and
analyzed using Quantitative Insights into
Microbial Ecology (QIIME, v.1.8.0) software
(38). Paired forward and reverse reads were
merged using fastq-join, demultiplexed,
and quality filtered (quality Phred score
Q . 20, fewer than three low-quality base
calls). Operational taxonomic units (OTUs)
were assigned using an open reference
picking strategy with Uclust (39) at 97%
identity against the 97% Greengenes
reference database (v13.8) (40). Alpha
diversity rarefaction curves for chao1 and
Shannon indexes were calculated in QIIME
using 10 iterations represented as mean 1
SD at different sequencing depths
(2,000–98,000). All downstream analyses
were performed using a rarefied OTU table
at 40,000 sequencing depth. A heatmap
displaying the OTUs represented at more
than 0.05% relative abundance in a
minimum of two samples was created using
the heatmap.2 function of the gplots
package in R (41). The dendrogram was
generated using Ward’s hierarchical
clustering with hclust algorithm in R on
Bray-Curtis dissimilarity matrix calculated
in QIIME.
Yadava, Pattaroni, Sichelstiel, et al.: Microbiota Sustains Chronic Pulmonary Inflammation
Preparation and Transfer of
Microbiota-enriched BAL Fluid
For preparing the microbiota-enriched BAL
fluid, BAL was performed three times using
1 ml of PBS per mouse. Five mice were
pooled per group. The BAL was then
centrifuged at 2,000 rpm for 10 minutes at
48 C, and the pellet containing mouse
cells was discarded. The BAL supernatant
was aliquoted into Biopur tubes and
centrifuged at 14,000 3 g for 10 minutes
at 48 C and washed with PBS at the same
speed. The final pellet was snap frozen in
liquid nitrogen and stored at 2808 C until
use. Before intranasal administration,
each aliquot was resuspended in 2 ml of
PBS and administered to recipient mice in
a volume of 100 ml per mouse
intranasally.
Neutralization of IL-17A
For neutralization of IL-17A, mice were
treated with 250 mg of anti–IL-17A
(clone 17F3) or the corresponding isotype
control antibody (clone MPOC-21) from
BioXCell (West Lebanon, NH). Antibodies
were administered intraperitoneally on
Days 6, 10, 13, 17, 20, and 24 after the first
challenge of LPS/elastase challenge.
Results
Murine Model of Chronic Pulmonary
Inflammation Mimics Key
Pathological Features of Human
Disease
To study the underlying pathogenic
mechanisms of COPD, we used a murine
model of chronic pulmonary inflammation
(Figure 1A). Mice were treated with a
combination of LPS and elastase over
4 weeks. Coadministration of LPS/elastase
resulted in an increase in compliance,
(P , 0.05) (Figure 1B) and a decrease in
the percentage of FEV0.1/FVC (P , 0.01)
(Figure 1C). The decline in lung function
was associated with emphysematous
changes in the lung parenchyma, quantified
by both mean linear intercept and
destruction index parameters (P , 0.001)
(Figures 1D–1F) in addition to
peribronchiolar and perivascular
infiltration of the lung tissue (Figure 1G).
There was also an enhanced cellular
infiltration in the BALF (P , 0.001)
(Figure 1H), with marked increase in
macrophages (P , 0.05), neutrophils
(P , 0.05), and lymphocytes (P , 0.001)
(Figure 1I). Notably, the emphysema and
inflammation persisted at least 2 months
after the cessation of LPS/elastase
treatment, (P , 0.001) (see Figures
E1A–E1C in the online supplement).
Although neutrophilia waned in the BAL,
the continued prevalence of hemosiderin-
laden macrophages (P , 0.01) was
indicative of chronic inflammation and
microhemorrhaging months after the final
challenge (Figure E1D) (42).
Chronic Pulmonary Inflammation Is
Associated with Lymphoid
Neogenesis, Enhanced Local
Antibody Responses, and IL-17A
Production
In line with previous data in human
disease and animal models of cigarette
smoke–induced emphysema (24–27), we
observed the formation of lymphoid
follicles (LFs), which could be categorized
as bronchiolar-associated lymphoid tissue
(P , 0.01), alveolar-associated lymphoid
tissue (P , 0.01), or vascular-associated
lymphoid tissue (VALT) (P , 0.001) based
on their localization (Figure 2A). We found
that VALT accounted for the majority of
LFs. Moreover, these LFs persisted even in
the absence of inflammatory stimuli in mice
that were analyzed 9 weeks after the last
challenge (P , 0.05) (Figure E1E). Overall,
there was an increase in the frequency
of germinal center B cells (P , 0.05)
(Figure 2B) as identified by Gl7 staining
and T follicular helper cells (P , 0.01)
(Figure 2C) as defined by the coexpression
of PD-1 (programmed cell death protein 1)
and Gl7 in the lungs of LPS/elastase–treated
mice. To assess whether the presence of LFs
correlated with an increase in the local
antibody response, we measured the
amount of the antibody isotypes IgG1,
IgA, and IgM in the BAL of PBS- or
LPS/elastase–treated mice. All isotypes were
increased in the BAL of LPS/elastase–
treated mice (P , 0.05) (Figure 2D). Of
note, no increase in the systemic levels of
any isotype was detected (Figure E2A).
We then investigated the reactivity of
these antibodies against collagen and elastin,
which constitute a component of the lung
tissue breakdown products. A significant
increase in specific responses against
collagen was detected for IgG1, (P , 0.01)
and IgA (P , 0.001) but not IgM isotypes
(Figures 2E–2G). A marked increase in
specific antibodies against elastin was
observed for IgA (P , 0.05), IgG1
977
 ORIGINAL ARTICLE

A B *
C **
PBS or

Compliance (cm H2O.s/ml)

LPS/elastase (L/E) 0.10 100
intranasally

FEV 0.1/FVC (%)

0.08
80

0.06
Week 1 2 3 4 5
60

Final Analysis PBS LPS/elastase PBS LPS/elastase

D E F
*** ****

Mean linear intercept(m)

PBS LPS/elastase 150 100

Destruction index
80
100
60

40
50
20

PBS LPS/elastase PBS LPS/elastase

G H I
PBS LPS/elastase 8
*** 6.5 PBS

Number in BAL(X105)
Total cells in BAL(X105)

* * LPS/elastase
4.5
6
**
2.5
4
0.5
2 0.2
0.1

PBS LPS/elastase Mac Neu Lym Eos

Figure 1. Murine model of chronic pulmonary inflammation mimics key pathological features of human disease. (A) BALB/c mice were administered
LPS (7 mg) and elastase (1.5 U) in a volume of 100 ml intranasally once a week for 4 weeks. Terminal analysis was performed 1 week after the last
challenge. (B) Lung compliance and (C) FEV0.1/FVC were measured by FlexiVent invasive airway mechanics system. (D) Paraformaldehyde-fixed
sections of lung were stained with hematoxylin and eosin, and emphysema and destruction of lungs were scored by (E) mean linear intercept and (F)
destruction index (n = 3–5). (G) Representative hematoxylin and eosin–stained lung sections showing perivascular and peribronchiolar inflammation
in LPS/elastase–treated mice. (H) The total cell and (I) differential counts in the bronchoalveolar alveolar lavage (BAL) were determined (n = 5). Error bars
represent SEM. Data representative of at least two independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Eos = eosinophils;
Lym = lymphocytes; Mac = macrophages; Neu = neutrophils; PBS = phosphate-buffered saline.

(P , 0.01), and IgM (P , 0.05) antibody
isotypes (Figures 2E–2G). We did not
detect an increase in systemic levels of
elastin-specific IgG1 and IgA responses,
although collagen-specific antibodies were
increased (Figures E2B and E2C). To
further characterize the autoimmune
aspect of the disease, we enumerated
IL-17–producing T cells (Figures 2H–2K). We
found that IL-17 production by both gd1
(P , 0.0001) (Figure 2H) and CD41 T cells
(P , 0.0001) was increased in the diseased
lungs (Figure 2I). IL-17A production was
also induced when sorted gd1 (Figure 2J)
and CD41 T cell (Figure 2K) populations
isolated from the lungs of LPS/elastase-
treated mice were restimulated with
collagen or elastin in vitro. Furthermore, we LPS/elastase over 4 weeks and assessed
found that the protein levels of IL-6 and microbiota diversity and composition by
IL-1b, cytokines that can enhance IL-17A Illumina 16S rDNA sequencing using a
responses, in addition to levels of IL-17A custom protocol. Indeed, amplifying
itself, were increased in the BALF of bacterial DNA from mouse BAL is
LPS/elastase–treated animals (Figures particularly challenging given the low
E3A–E3C) biomass and subsequent high risk of
contamination. For these reasons, we
The Pulmonary Microbiota Is Altered pooled BAL from four mice to increase the
upon the Induction of Chronic bacterial biomass, and we used stringent
Pulmonary Inflammation negative controls where PBS (as used for
It has been reported that humans with the BAL) was exposed to the same
COPD have an altered airway microbiota environment, tubing, and extraction kits as
(10, 13); thus, we investigated whether the BAL from mice. A negative control,
similar changes occurred in our mouse where we successfully found an amplified
model. We extracted bacterial DNA bacterial 16S rDNA fragment, was included
from BAL of mice treated with PBS or in the sequencing analyses to allow

978 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 9 | May 1 2016
ORIGINAL ARTICLE

A BALT VALT ALT 20 ****

No./section of the lung

PBS
LPS/elastase
15
**
10
**
5
100 μm 100 μm 100 μm
0
VALT BALT ALT

B GC B cells C Tfh cells

No.in the lung (X104)

No.in the lung (X105)

PBS LPS/elastase 4 * PBS LPS/elastase 4 ***

3 3
2 2
GL-7

GL-7
1 1
0 0
PD-1 PBS LPS/elastase PD-1 PBS LPS/elastase

D Total antibodies in BAL E Specific IgG1 F Specific IgA G Specific IgM

50 0.08 ** ** 0.05 *** 0.6 *
** PBS *
40 LPS/elastase 0.06 0.04
0.4
30 0.03
μg/ml

μg/ml
μg/ml

μg/ml
0.04
20 0.02
**** 0.2
0.02
10 *** 0.01
nd nd
0
IgG1 IgA IgM Collagen Elastin Collagen Elastin Collagen Elastin

H γδ+ T cells I
**** Media Collagen Elastin
PBS LPS/elastase 80
0.74% 2.48% 3.46%
%IL-17A+ of
γδ+ T cells

60
40
IL-17A
IL-17A

20

IFNγ PBS LPS/elastase γδ

J CD4+ T cells **** K Media Collagen Elastin

40
PBS LPS/elastase 0.091% 0.44% 0.36%
CD4+ T cells
%IL-17A+ of

30
20
IL-17A

IL-17A

10

IFNγ PBS LPS/elastase CD4

Figure 2. Chronic pulmonary inflammation is associated with lymphoid neogenesis, enhanced local antibody responses, and IL-17A production. (A)
Representative pictures of bronchial-associated lymphoid tissue (BALT), vessel-associated lymphoid tissue (VALT), and alveolar-associated lymphoid
tissue (ALT) in hematoxylin and eosin–stained lung sections and their quantification. Data pooled from two experiments (n = 6–8). (B) Germinal center (GC)
B cells and (C) follicular T helper (Tfh) cells from lungs were quantified by flow cytometry. Data pooled from two experiments. (D) Levels of IgG1, IgA,
and IgM in bronchoalveolar lavage (BAL) supernatant were measured by ELISA. Data pooled from two experiments (n = 8–9). Collagen- and elastin-
specific (E) IgG1, (F) IgA, and (G) IgM responses were quantified by ELISA. Data pooled from two experiments (n = 8–9). (H) Cells isolated from lungs and
airways of indicated groups were stimulated in vitro with phorbol myristate acetate, ionomycin, and monensin for 4 hours. IL-17 production by gd1 T cells
was quantified by flow cytometry. Data are representative of at least three independent experiments. (I) IL-17 production by CD41 T cells was quantified by

Yadava, Pattaroni, Sichelstiel, et al.: Microbiota Sustains Chronic Pulmonary Inflammation 979
 ORIGINAL ARTICLE

conclusions to be drawn from low-
abundance bacterial sequences. We found
that microbiota richness and diversity
were decreased in the pooled BAL of
LPS/elastase–treated mice, as indicated by
chao1 richness and Shannon diversity
rarefaction curves (Figure 3A). Of note, the
Shannon index diversity was higher in the
negative control than the LPS/elastase
group, potentially indicative of an increased
abundance of certain bacteria in the
LPS/elastase group, which is reflected by
the abundance component of the Shannon
diversity calculation. We next assessed
microbiota composition, with particular
attention to the negative control. We
observed that the most abundant
OTUs present in BAL samples, such
as family Comamonadaceae, family
Flavobacteriaceae, and family
Microbacteriaceae were also predominant
in the negative control, but their relative
abundance did not differ among BAL
samples (Figure 3B). Importantly,
Ward hierarchical clustering algorithm
using Bray-Curtis distance efficiently
distinguished and clustered BAL samples of
the treatments, as compared with the
negative control (Figure 3B). These data
show that a distinct microbiota was present
in healthy airways, diseased airways, and
the background control. To further
investigate the changes in microbiota
induced by LPS/elastase treatment, we
excluded contaminant OTUs and assessed
the differentially abundant genera across
the treatments. We identified four genera of
particular note: Prevotella, Pseudomonas,
Lactobacillus, and Chryseobacterium
(Figure 3C). Prevotella was present in the
BAL of the PBS control group but absent in
the BALF of LPS/elastase–treated mice.
On the contrary, relative abundance of
Pseudomonas, Lactobacillus, and
Chryseobacterium was increased in the
BAL of LPS/elastase–treated mice. Overall,
as reported for the human airway
microbiota, the airways of mice with this
chronic lung disease are characterized by a
distinct microbiota to those with healthy
airways.
Microbiota-derived Signals Amplify
Innate and Adaptive Inflammation in
Disease
To investigate whether microbiota-
dependent signals contribute to the
persistent inflammation, we compared
responses between SPF mice and mice
depleted of a microbiota using an antibiotic
treatment regime. Before the start of the
experiment mice were given 10% Baytril in
drinking water for 2 weeks followed by
treatment with Co-Amoxi-Mepha for
another 2 weeks. This treatment decreased
the microbial load by approximately 6 log,
(P , 0.0001) (Figure E4). During the course
of the experiment mice were maintained on
Co-Amoxi-Mepha treatment. We found
that antibiotic-treated mice showed no
change in FEV0.1/FVC (Figure 4A) but a
significant reduction in lung compliance
(P , 0.05) (Figure 4B). This was linked to a
decrease in inflammation in the lungs and
airways (Figure 4C), decreased total cell
counts (P , 0.01) (Figure 4D), and
neutrophil (P , 0.01) and interstitial
macrophage (P , 0.01) infiltration in the
BAL (Figures 4E and 4F). Macrophage
subsets were identified as described in
Figure E5A. Antibiotic treatment also
resulted in a reduction in the frequency of
VALT (P , 0.05) (Figure 4G). Although
the total antibody levels of IgG1 and IgM in
the BAL were unaffected in antibiotic-
treated mice (Figure 4H), collagen-specific
IgG1 responses were significantly reduced
(P , 0.01) (Figure 4I). Comparatively, total
IgA responses were reduced (P , 0.01)
(Figure 4H), and, in line with this, collagen-
and elastin-specific IgA responses were
reduced (P , 0.05) (Figure 4J).
Microbiota Enhances Production of
IL-17A by T Cells
Microbiota-dependent signals have been
shown to amplify autoimmune diseases
in an IL-17–dependent manner (34–36).
Both autoimmune mechanisms and IL-17
(31, 46–54) are implicated in the pathogenesis
of COPD. Antibiotic-treated mice showed
a reduced IL-17A response (P , 0.05)
(Figure 5A). Upon investigation of the cellular
sources of IL-17A, we found no difference
in the composition of cells producing IL-
17A (Figure 5B) using the gating strategy
shown in Figure E5B, and both CD41 and
gd1 T cells accounted for the majority
of IL-17A producers. However, the
frequencies of IL-17A1 within CD41
T cells and the total number of IL-17A1
CD41 T cells were markedly reduced upon
antibiotic treatment (P , 0.01), whereas
only a minor reduction was seen for
gd1 T cells, which was not statistically
significant (Figures 5C and 5D). Similar
to antibiotic-treated mice, axenic mice
exhibited reduced recruitment of
neutrophils and macrophages in the
airways (P , 0.05, P , 0.01, respectively)
(Figures E6A–E6D) and a specific reduction
in IL-17A-producing gd1 T cells (P , 0.05)
(Figure E6E).
In an attempt to assess whether the
altered microbiota in disease could directly
enhance IL-17A responses, we performed a
transfer experiment. Antibiotic-treated mice
were challenged with LPS/elastase on Days
0, 7, and 21. Concurrently, they were given a
solution that was enriched for the airway
microbiota of PBS- or LPS/elastase–treated
mice intranasally. We found that the
transfer of this microbiota-enriched
solution from LPS/elastase–treated mice
resulted in an increase in the IL-
17A–producing cells in the lungs of
recipient mice (Figure 5E). In particular,
IL-17A–producing CD41 T cells
(Figure 5F) were increased (P , 0.05),
whereas gd1 T cells (Figure 5G) were
unaffected. Performing the same
experiment using axenic recipients, we
observed an increase in the proportion of
IL-17A–producing CD41 and gd1 T cells
(P , 0.05) (Figure E6F).
IL-17A Kinetics Correlates with
Disease Severity, and Targeted
Neutralization of IL-17A Ameliorates
Disease
We next sought to determine the kinetics of
IL-17A production during the induction of
disease. We modified our existing model to
recapitulate different levels of severity

Figure 2. (Continued). flow cytometry in cells isolated from lungs and airways of indicated groups stimulated in vitro with phorbol myristate acetate,
ionomycin, and monensin for 4 hours. Data are representative of at least three independent experiments. (J) Fluorescence-activated cell sorted gd1 T cells
from pooled lungs and airway samples of five mice treated with LPS/elastase were stimulated with bone marrow–derived dendritic cells (BMDCs)
alone (media), or BMDCs pulsed with collagen or elastin. (K) CD41 T cells were sorted by flow cytometry from pooled lung and airway samples of
five mice treated with LPS/elastase. Sorted CD41 T cells were then stimulated with BMDCs alone (media) or BMDCs pulsed with collagen or elastin.
*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. PBS = phosphate-buffered saline; PD-1 = programmed cell death protein 1.

980 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 9 | May 1 2016
ORIGINAL ARTICLE

A chao1 (Richness) Shannon (Diversity)

2000 6.3

6.2

6.1
1500
6.0

Shannon
5.9
chao1

1000
5.8

PBS (4 pooled BALF) 5.7 PBS (4 pooled BALF)

PBS (4 pooled BALF) PBS (4 pooled BALF)
500 5.6
LPS/Elastase (4 pooled BALF) LPS/Elastase (4 pooled BALF)
LPS/Elastase (4 pooled BALF) 5.5 LPS/Elastase (4 pooled BALF)
Negative control 5 Negative control
0 0

00

10 0
14 0
18 0
22 0
26 0
30 0
34 0
38 0
42 0
46 0
50 0
54 0
58 0
62 0
66 0
70 0
74 0
78 0
82 0
86 0
90 0
94 0
98 0
0
00

10 0
14 0
18 0
22 0
26 0
30 0
34 0
38 0
42 0
46 0
50 0
54 0
58 0
62 0
66 0
70 0
74 0
78 0
82 0
86 0
90 0
94 0
98 0
0

0
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
0
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00

20
60
20
60

N seqs/sample N seqs/sample

B hclust (ward.D), Bray-Curtis distance Relative abundance %

0 10 20 30
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__;g__
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__ACK−M1;g__
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Actinomycetaceae;g__N09
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Corynebacteriaceae;g__Corynebacterium
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Geodermatophilaceae;g__
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Microbacteriaceae;g__
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Microbacteriaceae;g__Candidatus Rhodoluna
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Microbacteriaceae;g__Clavibacter
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Microbacteriaceae;g__Microbacterium
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Micrococcaceae;Other
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Micrococcaceae;g__Micrococcus
k__Bacteria;p__Actinobacteria;c__Actinobacteria;o__Actinomycetales;f__Propionibacteriaceae;g__Propionibacterium
k__Bacteria;p__Bacteroidetes;c__Bacteroidia;o__Bacteroidales;f__Porphyromonadaceae;g__Paludibacter
k__Bacteria;p__Bacteroidetes;c__Bacteroidia;o__Bacteroidales;f__Prevotellaceae;g__
k__Bacteria;p__Bacteroidetes;c__Bacteroidia;o__Bacteroidales;f__Prevotellaceae;g__Prevotella
k__Bacteria;p__Bacteroidetes;c__Bacteroidia;o__Bacteroidales;f__Rikenellaceae;g__
k__Bacteria;p__Bacteroidetes;c__Bacteroidia;o__Bacteroidales;f__S24−7;g__
k__Bacteria;p__Bacteroidetes;c__Bacteroidia;o__Bacteroidales;f__[Odoribacteraceae];g__Odoribacter
k__Bacteria;p__Bacteroidetes;c__Cytophagia;o__Cytophagales;f__Cytophagaceae;g__
k__Bacteria;p__Bacteroidetes;c__Flavobacteriia;o__Flavobacteriales;f__Cryomorphaceae;g__
k__Bacteria;p__Bacteroidetes;c__Flavobacteriia;o__Flavobacteriales;f__Cryomorphaceae;g__Fluviicola
k__Bacteria;p__Bacteroidetes;c__Flavobacteriia;o__Flavobacteriales;f__Flavobacteriaceae;g__Flavobacterium
k__Bacteria;p__Bacteroidetes;c__Flavobacteriia;o__Flavobacteriales;f__[Weeksellaceae];Other
k__Bacteria;p__Bacteroidetes;c__Flavobacteriia;o__Flavobacteriales;f__[Weeksellaceae];g__Chryseobacterium
k__Bacteria;p__Bacteroidetes;c__Flavobacteriia;o__Flavobacteriales;f__[Weeksellaceae];g__Cloacibacterium
k__Bacteria;p__Bacteroidetes;c__[Saprospirae];o__[Saprospirales];f__Chitinophagaceae;g__Niabella
k__Bacteria;p__Cyanobacteria;c__Chloroplast;o__Cryptophyta;f__;g__
k__Bacteria;p__Cyanobacteria;c__Chloroplast;o__Stramenopiles;f__;g__
k__Bacteria;p__Cyanobacteria;c__Chloroplast;o__Streptophyta;f__;g__
k__Bacteria;p__Deferribacteres;c__Deferribacteres;o__Deferribacterales;f__Deferribacteraceae;g__Mucispirillum
k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;f__Staphylococcaceae;g__Staphylococcus
k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;f__Lactobacillaceae;g__Lactobacillus
k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;f__Streptococcaceae;g__Streptococcus
k__Bacteria;p__Firmicutes;c__Bacilli;o__Turicibacterales;f__Turicibacteraceae;g__Turicibacter
k__Bacteria;p__Firmicutes;c__Clostridia;o__Clostridiales;f__;g__
k__Bacteria;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Oscillospira
k__Bacteria;p__Firmicutes;c__Erysipelotrichi;o__Erysipelotrichales;f__Erysipelotrichaceae;g__[Eubacterium]
k__Bacteria;p__GN02;c__3BR−5F;o__;f__;g__
k__Bacteria;p__GN02;c__GKS2−174;o__;f__;g__
k__Bacteria;p__OD1;c__SM2F11;o__;f__;g__
k__Bacteria;p__OD1;c__ZB2;o__;f__;g__
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__;f__;g__
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Caulobacterales;f__Caulobacteraceae;g__
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Caulobacterales;f__Caulobacteraceae;g__Caulobacter
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Rhizobiales;f__Hyphomicrobiaceae;g__Hyphomicrobium
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Rhizobiales;f__Methylobacteriaceae;g__Methylobacterium
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Rhizobiales;f__Phyllobacteriaceae;Other
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Rhodobacterales;f__Rhodobacteraceae;g__Rhodobacter
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Rickettsiales;f__Rickettsiaceae;g__
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Sphingomonadales;f__;g__
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Sphingomonadales;f__Erythrobacteraceae;g__
k__Bacteria;p__Proteobacteria;c__Alphaproteobacteria;o__Sphingomonadales;f__Sphingomonadaceae;g__
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__;f__;g__
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Burkholderiales;f__Burkholderiaceae;g__Burkholderia
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Burkholderiales;f__Comamonadaceae;Other
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Burkholderiales;f__Comamonadaceae;g__
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Burkholderiales;f__Comamonadaceae;g__Delftia
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Burkholderiales;f__Oxalobacteraceae;g__
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Burkholderiales;f__Oxalobacteraceae;g__Cupriavidus
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Burkholderiales;f__Oxalobacteraceae;g__Polynucleobacter
k__Bacteria;p__Proteobacteria;c__Betaproteobacteria;o__Procabacteriales;f__Procabacteriaceae;g__
k__Bacteria;p__Proteobacteria;c__Deltaproteobacteria;o__Bdellovibrionales;f__Bdellovibrionaceae;g__Bdellovibrio
k__Bacteria;p__Proteobacteria;c__Deltaproteobacteria;o__MIZ46;f__;g__
k__Bacteria;p__Proteobacteria;c__Deltaproteobacteria;o__Myxococcales;f__;g__
k__Bacteria;p__Proteobacteria;c__Deltaproteobacteria;o__Spirobacillales;f__;g__
k__Bacteria;p__Proteobacteria;c__Epsilonproteobacteria;o__Campylobacterales;f__Campylobacteraceae;g__Sulfurospirillum
k__Bacteria;p__Proteobacteria;c__Gammaproteobacteria;o__Alteromonadales;f__Alteromonadaceae;g__Cellvibrio
k__Bacteria;p__Proteobacteria;c__Gammaproteobacteria;o__Alteromonadales;f__[Chromatiaceae];g__Rheinheimera
k__Bacteria;p__Proteobacteria;c__Gammaproteobacteria;o__Legionellales;f__Legionellaceae;g__Legionella
k__Bacteria;p__Proteobacteria;c__Gammaproteobacteria;o__Pseudomonadales;f__Moraxellaceae;g__Acinetobacter
k__Bacteria;p__Proteobacteria;c__Gammaproteobacteria;o__Pseudomonadales;f__Pseudomonadaceae;g__Pseudomonas
k__Bacteria;p__SR1;c__;o__;f__;g__
k__Bacteria;p__TM7;c__;o__;f__;g__
k__Bacteria;p__TM7;c__TM7−1;o__;f__;g__
k__Bacteria;p__TM7;c__TM7−3;o__EW055;f__;g__
k__Bacteria;p__[Thermi];c__Deinococci;o__Deinococcales;f__Deinococcaceae;g__CM44
Negative LPS/Elastase PBS
control

C g_Prevotella g_Pseudomonas g_Lactobacillus g_Chryseobacterium

Relative abundance (%)

0.003 0.025 0.08 0.020

0.020
0.06 0.015
0.002
0.015
0.04 0.010
0.010
0.001
0.02 0.005
0.005

0.000 0.000 0.00 0.000

PBS LPS/Elastase PBS LPS/Elastase PBS LPS/Elastase PBS LPS/Elastase

Figure 3. The pulmonary microbiota is altered upon the induction of chronic pulmonary inflammation. Eight mice per group were treated with phosphate-
buffered saline (PBS) or LPS/elastase for 4 weeks, and DNA was extracted from four pooled bronchoalveolar lavage fluid (BALF) samples per group,
resulting in two sample sets per group for Illumina MiSeq 16S rDNA sequencing. (A) Richness (chao1) and diversity (Shannon) rarefaction curves of the
16S rDNA library constructed from pooled BALF of PBS- or LPS/elastase–treated mice. (B) Heat map showing the relative abundance of major bacterial
genera (operational taxonomic units . 0.05% of relative abundance in minimum two samples) in pooled BALF of PBS- or LPS/elastase–treated mice and
negative control. Dendrogram was drawn from Ward hierarchical clustering algorithm using Bray-Curtis dissimilarity matrix. (C) Bar graphs representing
the relative abundance of specific genera showing differences between the BALF of PBS- and LPS/elastase–treated mice.

Yadava, Pattaroni, Sichelstiel, et al.: Microbiota Sustains Chronic Pulmonary Inflammation 981
 ORIGINAL ARTICLE

A B C
Water Antibiotic
110 0.09
FEV0.1/FVC(%)

(mL/cmH2O)
100

Compliance
0.08
90
0.07
80
70 0.06
60
Water Antibiotic Water Antibiotic 100 m 100 m

D ** E Water Antibiotic **
1.5 2.0

Neutrophils in BAL
Total cells in BAL

1.2 1.5
(X106)

(X105)
0.9
1.0
0.6
Ly6g

0.3 0.5

Water Antibiotic Ly6C Water Antibiotic

F Water Antibiotic
4 **
4
aM in BAL(X105)

iM in BAL(X104)
3 3
2 2
CD11c

1 1

F480 Water Antibiotic Water Antibiotic

G H Total antibodies in BAL I Specific IgG1 J Specific IgA

25 * Water 50
** 0.04 0.04 *
20 Antibiotic 40 ** **
No./section

0.03 0.03
of the lung

g/ml
g/ml

g/ml

15 30
0.02 0.02
10 20
5 10 0.01 0.01

VALT BALT ALT IgG1 IgA IgM Collagen Elastin Collagen Elastin
Figure 4. Microbiota-derived signals amplify innate and adaptive inflammation in disease. Mice harboring a specific pathogen–free microflora (water)
or mice depleted of microbiota by antibiotic treatment were subjected to LPS/elastase treatment over 4 weeks. Analysis was performed a week after last
challenge. (A) Lung compliance and (B) FEV/FVC were measured using FlexiVent invasive airway mechanics system. Data are pooled from two
experiments. (C) Representative slides of hematoxylin and eosin–stained lung sections showing peribronchiolar and perivascular inflammation in water-
and antibiotic-treated groups. (D) Number of cells infiltrating the bronchoalveolar lavage (BAL) was determined. (E) Neutrophils and (F) alveolar (aMF) and
interstitial (iMF) macrophages in the BAL were enumerated by flow cytometry. Data are representative of two independent experiments. (G) Vascular-
associated lymphoid tissues (VALT), alveolar-associated lymphoid tissues (ALT), or bronchial-associated lymphoid tissues (BALT) were quantified in
hematoxylin and eosin–stained sections of lungs. Error bars represent SEM (n = 4). (H) Levels of IgG1, IgA, and IgM in BAL supernatant were measured by
ELISA. Collagen- and elastin-specific (I) IgG1 and (J) IgA responses were quantified by ELISA. Error bars represent SEM. Data pooled from two
experiments (n = 9). *P < 0.05, **P < 0.01.

similar to COPD in humans. We were able T cells and CD41 T cells increasingly (Figure 6F). We found that neutralizing
to distinguish mild and severe disease based accumulated in the lungs with repeated IL-17A did not improve FEV0.1/FVC but
on the decline in lung function seen at LPS/elastase challenges (P , 0.01) (Figures reduced lung compliance significantly
different time points in the disease model 6D and 6E, Figure E7). Based on the (P , 0.05) (Figures 6G and 6H) in addition
(Figures 6A and 6B). We found that the kinetics observed in mild versus severe to the total cellular and neutrophilic
kinetics of IL-17A production followed disease, we performed a targeted infiltration into the BAL (P , 0.01)
the progression of disease (Figure 6C). neutralization of IL-17A starting 1 day (Figures 6I and 6J). Neutralizing IL-17A
Moreover, both IL-17–producing gd1 before the second LPS/elastase challenge also reduced the total frequency of

982 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 9 | May 1 2016
ORIGINAL ARTICLE

A B

IL-17A+ cells (X106)

* Sources of IL-17A
Gated on live cells 5.0
Water Antibiotic T cells
Water Antibiotic 4.0
3.98% 5.06% CD3- cells
3.0 14.33% 26.69 15.88%
31.05 CD3+CD4–,
2.0 % 0.56% % 0.57% CD8–,– cells
IL-17A

1.0 CD8+ T cells

50.09% 51.81% CD4+ T cells
IFN Water Antibiotic

C ** D
Gated on CD4+ T cells Gated on + T cells 2.0
1.5

T cells i(X106)
IL-17A+ CD4+
T cells (X106)

IL-17A+ +
Water Antibiotic Water Antibiotic 1.5
1.0
1.0
0.5
IL-17A

IL-17A
0.5

IFN Water Antibiotic IFN Water Antibiotic

E F G
Gated on live cells 1.5 * 1.5 1.5
*

T cells i(×105)
IL-17A+ CD4+

PBS LPS/Elastase

IL-17A+ +
cells (X106)
cells (X106)

microbiome Microbiome 1.0 1.0 1.0

IL-17A+

0.5 0.5 0.5

IL-17A

PBS LPS/elastase PBS LPS/elastase PBS LPS/elastase

IFN
Source of microbiome Source of microbiome Source of microbiome
Figure 5. Microbiota enhances production of IL-17A by CD41 and gd1 T cells. Cells isolated from lungs and airways of control (water) or antibiotic were
stimulated in vitro with phorbol myristate acetate, ionomycin, and monensin for 4 hours. (A) Ex vivo stimulated cells were surface stained, fixed, and
permeabilized for detection of cytokines IFN-g and IL-17A. (B) Cellular sources of IL-17A were analyzed by flow cytometry. Pie charts represent the
composition based on mean values from four to five mice per group. (C) Total number of CD41 T cells producing IL-17A or (D) gd1 T cells was determined
by flow cytometry. Data are representative of two independent experiments. (E) Antibiotic-treated mice were challenged with LPS/elastase on Day 0, 7,
and 21. Every day starting at Day 12 until Day 20 and then on Days 22 to 24, they received microbiota (sequenced in Figure 3) isolated from either
phosphate-buffered saline (PBS)-challenged specific pathogen–free (SPF) mice or LPS/elastase-challenged SPF mice in 100 ml intranasally. On Day 25,
cells isolated from the lungs of the recolonized antibiotic treated recipients were restimulated with phorbol myristate acetate ionomycin, and IL-17
production was analyzed by flow cytometry. Total number of IL-17A–producing (F) CD41 T cells and (G) gd1 T cells were also determined by flow
cytometry. *P < 0.05, **P < 0.01.

lymphoid follicles (P , 0.05) without
specifically affecting any particular subtype
(VALT, alveolar-associated lymphoid tissue,
or bronchiolar-associated lymphoid tissue)
(Figure 6K). This was also associated with a
reduction in total IgG1 (P , 0.05), whereas
IgA and IgM were increased (P , 0.05,
P , 0.01, respectively) (Figure 6L). Overall,
targeting the IL-17A pathway was effective at
ameliorating inflammation and uncoupling
the autoimmune component of the disease.
Discussion
Alterations in the gut microbiome have been
extensively documented for chronic diseases
such as inflammatory bowel disease (19, 20). changes seen in microbial composition
Although the gut microbiota can impact in the airways of mice as in humans.
systemic immune responses, there is Specifically, we found that the overall
an emerging paradigm supporting the diversity of the pulmonary microbiome
importance of host–microbe interactions was reduced, Prevotella was absent, and
localized within a single habitat, such as the bacterial genera Pseudomonas and
that found in the lung and skin (55, 56). Lactobacillus were increased in abundance
Several studies have shown that the in the LPS/elastase–treated animals as
composition of the airway microbiota in compared with control animals (43). This
healthy lungs differs from lungs of patients correlates with studies in humans showing
with COPD, although as yet whether that Prevotella is a common colonizer of the
disease creates a habitat for the bacteria or airways of healthy subjects when compared
whether the bacteria cause disease hasn’t with patients with asthma or COPD (13,
been delineated (10, 13, 14). Our findings 43). Consistently Pseudomonas is increased
implicate a functional role for the in patients with asthma and COPD (43),
microbiota in COPD. It is noteworthy that and its presence correlates with COPD
there were many similarities in the types of severity (44). Similarly, the relative

Yadava, Pattaroni, Sichelstiel, et al.: Microbiota Sustains Chronic Pulmonary Inflammation 983
 ORIGINAL ARTICLE

0.09 100

FEV 0.1/FVC (%)

A PBS or LPS/Elastase B * PBS

(cm H2O.s/ml)
**

Compliance
0.08 LPS/Elastase
80
0.07
0.06 60
Week 1 2 3 4 5 0.05
1 2 3 4 1 2 3 4
Analysis Analysis Analysis Analysis No. of LPS/Elastase No. of LPS/Elastase
challenges challenges

C D E

IL-17A+ CD4+ cells

No. of LPS/Elastase challenges 107 106

IL-17A+  cells

** 106 ** **
** **

IL-17A+ cells
**
1X 2X 3X 4X 106 10 5
**
* 10 5
* ***
IL-17A

105 * ** *
104 104

RORt 1 2 3 4 1 2 3 4 1 2 3 4
No. of LPS/Elastase No. of LPS/Elastase No. of LPS/Elastase
challenges challenges challenges

F PBS or LPS/Elastase G H *
Analysis
100 0.065
FEV0.1/FVC (%)

(mL/cmH2O)
Compliance
80
0.060
60
0 6 7 10 13 14 17 20 21 24 28 0.055
40
20 0.050

IL-17A or isotype control antibody treatment Isotype IL-17A Isotype IL-17A

I J K L
* * * * Isotype
Cells in BAL (X105)

30
Neutrophils (X105)

No./section of lung

4.5 1.5 50 IL-17A

4.0 g/ml
40
1.0 20
3.5 30 *
3.0 10 20
0.5 **
2.5 10
Isotype IL-17A Isotype IL-17A Isotype IL-17A IgG1 IgA IgM
Figure 6. IL-17A kinetics correlate with disease severity, and targeted neutralization of IL-17A ameliorates disease. (A) BALB/c mice were challenged with
one, two, three, or four doses of LPS/elastase. Analysis was performed 1 week after the indicated number of challenges. (B) Invasive lung function
measurement was performed using the FlexiVent system. (C) The kinetics of RORgt-expressing and IL-17A–producing ex vivo stimulated cells from lungs
and airways of mice challenged with one, two, three, or four times phosphate-buffered saline (PBS) or LPS/elastase (n = 4–5 per time point per group).
The kinetics of IL-17A–producing (D) CD41 and (E) gd1 T cells in mice challenged with one, two, three, or four times PBS or LPS/elastase (n = 4–5 per time
point per group). (F) Targeted neutralization of IL-17A based on its production in mild to severe disease. Mice were challenged with LPS/elastase
over 4 weeks and administered anti–IL-17A antibody or isotype control at indicated time points. (G) Lung compliance and (H) FEV/FVC were measured
using the FlexiVent invasive airway mechanics system. (I) Number of cells recovered in the bronchoalveolar lavage (BAL) was determined. (J) Neutrophils
in the BAL were enumerated by differential cell counts. Data are representative of at least two independent experiments. (K) Vascular-associated,
alveolar-associated, or bronchial-associated lymphoid follicles were quantified in hematoxylin and eosin sections of lungs. Error bars represent SEM
(n = 4–5). (L) Levels of IgG1, IgA, and IgM in BAL supernatant were measured by ELISA. Data pooled from two experiments (n = 9). *P < 0.05, **P < 0.01,
***P < 0.001. RORgt = retinoid-related orphan receptor gt.

abundance of Lactobacillus is also increased A key question still facing the field is pressure facilitating colonization by specific
in patients with asthma and COPD (43), where does the airway microbiota originate? bacteria. Considering our experiments were
and an additional study has linked it to (57). Recent studies suggest the pulmonary performed under controlled conditions
COPD severity (10). Chryseobacterium microbiome originates from the dispersal of with mice housed in isolated and filtered
levels were also increased in the BAL of microbes from the nasal and oral cavities cages, our results support the concept
LPS/elastase–treated mice. This did not (58, 59). It is possible that this “seeding” that inflammatory conditions allow the
correlate with any published data in population is altered in patients with outgrowth of bacteria that are part of the
patients with COPD, but studies have COPD, and this subsequently affects the steady-state “healthy” microbiota.
shown that Chryseobacterium infection in microbiome in the lungs. Alternatively, Increased levels of IL-17A have been
individuals with cystic fibrosis correlates the state of chronic inflammation in the implicated in stable COPD (52), and a
with Pseudomonas (45). diseased lungs may provide a selection recent study has shown that the number of

984 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 9 | May 1 2016
ORIGINAL ARTICLE
CD41 IL-17A1 cells in the small airways
positively correlates with airflow limitation
(46). Tracking the development of disease
progression in our model showed that
IL-17A production increased with disease
severity. Moreover, we observed temporal
changes in the cellular sources of IL-17A
(Figure E7). We found that gd1 T cells
accounted for the majority of the few
IL-17A–producing cells in the healthy and
mildly diseased lungs. In severe disease
there was an expansion of both
IL-17A–producing CD41 and gd1 T cells.
Signals from commensal bacteria have been
shown to enhance the expansion of both
gd1 T cells and Th17 cells (34, 60–62) and
exacerbate IL-17–mediated inflammation
in experimental autoimmune encephalitis
and arthritis (34–36). In particular, specific
bacteria can induce Th17 responses, among
which segmented filamentous bacteria have
been extensively studied (34, 36). Although
segmented filamentous bacteria were not
detected in the lung in our study, it is likely
that the outgrowth of other lung bacteria
resident in the diseased lungs could also
promote IL-17 responses.
In the current study, we attempted to
transfer the microbiota from the BAL of
either PBS- or LPS/elastase–treated mice
into the airways of mice depleted (with
antibiotics) or devoid (germ-free) of a
microbiota. This transfer consisted of a
microbial pellet that had been isolated
(excluding host cells and soluble
components in suspension) from the BAL
performed on mice; however, in addition to
the enriched microbial components of this
pellet, acellular material from the lungs
could also have been enriched. Transfer of
the microbiota-enriched solution derived
from mice that had previously been treated
with LPS/elastase increased the production
of IL-17 in recipient mice, as compared
with the solution derived from mice
previously treated with PBS. Whether this
experimental approach led to a sufficiently
large increase in IL-17 levels to influence
lung function remains to be determined.
There are several technical limitations
to the experimental approach of transferring
the airway microbiota. For instance, it is
unclear whether viable bacteria are required
to see this effect. Another important caveat
is we are unable to distinguish between the
influence of the microbes and that of other
potential constituents of the processed
BALF, for example, cellular debris. With the
current state of the art and the very low
Yadava, Pattaroni, Sichelstiel, et al.: Microbiota Sustains Chronic Pulmonary Inflammation
bacterial biomass in the airways of mice, we
are limited with feasible approaches to
address this point. As such, from this study
we can only conclude the microbiota
changes in the airways of mice treated with
LPS/elastase, similar to that reported in
humans; depletion (antibiotics) or absence
(axenic/germ-free mice) of a microbiota
ameliorates disease; and transfer of a
microbially enriched fraction of the BAL
from diseased animals increases the disease-
causing IL-17 inflammatory pathway in
recipient mice. A further possibility is that
the role of the microbiota in promoting
disease in our model is due to the gut
microbiota; indeed, we have previously
reported a gut–lung axis, with mice fed a
high-fiber diet being protected against
allergic airway inflammation (63).
However, in the current study we expect
the airway microbiota to play a more
fundamental role, as we did not find
changes in the gut microbiota in our model
(data not shown). It remains possible that
certain changes in the gut microbiota could
impact the development of COPD, and
future in-depth metagenomic approaches
could provide the answer.
Concurrent with the development of
the IL-17 response in our model, we
found an autoimmune component of the
inflammation evolved. It is tempting to
speculate that the microbially enhanced
IL-17 response promotes an autoimmune
component, which further perpetuates
the disease. The increased availability of
self-antigens due to the destruction of lung
tissue might also allow the expansion of
autoreactive T cells. Further supporting
an autoimmune component to disease
pathogenesis, lymphoid follicles have
consistently been observed in animal
models and in humans with COPD
(24–27). Recently, Bracke and colleagues
demonstrated a pathogenic role of LFs in
disease (26). In this study they abrogated
LF formation by targeting CXCL13 and
observed a decrease in the local antibody
response with a concomitant improvement
in specific aspects of disease such as
inflammation and alveolar wall destruction.
Microbial colonization and infections are
more frequent in patients with very severe
disease. From an evolutionary standpoint,
LFs in chronic lung diseases might arise to
combat infections by enhancing the local
immune response. However, self-antigens
derived from the collateral damage of the
lung tissue might also be presented within
LFs and result in the production of
autoantibodies (22, 64). Although sequence
analysis of B cell clones isolated from LFs in
COPD lungs has indicated the presence of
an oligoclonal, antigen-specific humoral
response (25), there are conflicting reports
regarding the production of self-reactive
antibodies (28–32). It is important to note
that most of these studies assess systemic
antibody levels, which may not be
representative of responses within the lung.
Another caveat in these studies is the
identification of the relevant autoantigen,
which may vary considerably among
patients. In our model and in the cigarette
smoke–induced model of emphysema, LF
formation is associated with an increase in
local antibody levels in the BAL, whereas
systemic antibodies were unaffected (25,
26). In line with these data, the deposition
of IgG complexes has been demonstrated
in lungs of patients with severe COPD,
although their antigenic specificity is
unknown (30, 65). A limitation of our study
is that we investigated self-reactive
antibodies to only two self-antigens,
collagen and elastin, which constitute only
a portion of the breakdown products of the
lung tissue. In the future it would be
valuable to comprehensively characterize
the repertoire of the self-reactive antibodies
in patients with COPD and investigate
whether specific antibody signatures are
correlated with disease severity. This could
distinguish a subset of patients with COPD
in whom disease pathogenesis has a clear
association with autoimmunity and
ultimately identify patients who would
benefit the most from treatments targeting
B cells (66, 67).
The beneficial versus detrimental role
of local antibody responses could be
determined by different antigen specificities
and associated isotypes. It has been reported
that compared with patients with COPD,
healthy smokers exhibit a preferential
switching to IgA (68). IgA responses at
mucosal surfaces are associated with
protective immunity, and blunted IgA
responses in COPD can be linked to
increased susceptibility to infections (69).
Conversely, self-reactive antibodies, in
particular IgG1 isotypes, are more prevalent
in severe disease and implicated in
pathogenesis (30, 65, 68). In our study,
depleting the microbiota had a profound
effect on the total IgA levels, suggesting that
a proportion of these antibody responses
may be directed against the microbiota.
985
 ORIGINAL ARTICLE

However, the absence of a microbiota also
reduced the levels of self-reactive IgA and
IgG1 antibodies. Whether this is due to the
cross-reactivity between microbial and
self-antigens or an indirect effect by the
induction of IL-17, as seen in a model of
autoimmune arthritis (34), is unknown.
Our data suggest that microbial cues could
impact not only the specificity but also the
isotype of antibody responses generated
locally.
Our data implicate the airway
microbiota as a key player in the promotion
of local IL-17A responses and, consequently,
the development of chronic lung disease.
This pathway impacts both innate and
adaptive inflammatory processes and
regulates lymphoid neogenesis and the
expansion of autoreactive T and B cells. The
microbiota-enhanced local autoimmunity
likely contributes to disease chronicity,
although autoimmunity only represents one
pathogenic mechanism linked with
COPD. The characterization of dysbiotic
microbial communities and their
associated inflammatory signature could
pave the way for personalized medicine
to disrupt the chronic cycle of tissue
damage and the heterogeneity seen in
COPD. n
Author disclosures are available with the text
of this article at www.atsjournals.org.

References
1. Rosenbaum L, Lamas D. Facing a “slow-motion disaster”: the UN
meeting on noncommunicable diseases. N Engl J Med 2011;365:
2345–2348.
2. Halbert RJ, Natoli JL, Gano A, Badamgarav E, Buist AS, Mannino DM.
Global burden of COPD: systematic review and meta-analysis.
Eur Respir J 2006;28:523–532.
3. Fletcher C, Peto R. The natural history of chronic airflow obstruction.
BMJ 1977;1:1645–1648.
4. Saetta M, Finkelstein R, Cosio MG. Morphological and cellular basis for
airflow limitation in smokers. Eur Respir J 1994;7:1505–1515.
5. Tuder RM, Petrache I. Pathogenesis of chronic obstructive pulmonary
disease. J Clin Invest 2012;122:2749–2755.
6. Finkelstein R, Fraser RS, Ghezzo H, Cosio MG. Alveolar inflammation
and its relation to emphysema in smokers. Am J Respir Crit Care Med
1995;152:1666–1672.
7. Barnes PJ. Mediators of chronic obstructive pulmonary disease.
Pharmacol Rev 2004;56:515–548.
8. Willemse BW, ten Hacken NH, Rutgers B, Lesman-Leegte IG, Postma
DS, Timens W. Effect of 1-year smoking cessation on airway
inflammation in COPD and asymptomatic smokers. Eur Respir J 2005;
26:835–845.
9. Zakharkina T, Heinzel E, Koczulla RA, Greulich T, Rentz K, Pauling JK,
Baumbach J, Herrmann M, Grünewald C, Dienemann H, et al. Analysis
of the airway microbiota of healthy individuals and patients with
chronic obstructive pulmonary disease by T-RFLP and clone
sequencing. Plos One 2013;8:e68302.
10. Sze MA, Dimitriu PA, Hayashi S, Elliott WM, McDonough JE, Gosselink
JV, Cooper J, Sin DD, Mohn WW, Hogg JC. The lung tissue
microbiome in chronic obstructive pulmonary disease. Am J Respir
Crit Care Med 2012;185:1073–1080.
11. Han MK, Huang YJ, Lipuma JJ, Boushey HA, Boucher RC, Cookson
WO, Curtis JL, Erb-Downward J, Lynch SV, Sethi S, et al.
Significance of the microbiome in obstructive lung disease. Thorax
2012;67:456–463.
12. Erb-Downward JR, Huffnagle GB, Martinez FJ. The microbiota in
respiratory disease. Am J Respir Crit Care Med 2012;185:1037–1038.
13. Erb-Downward JR, Thompson DL, Han MK, Freeman CM, McCloskey
L, Schmidt LA, Young VB, Toews GB, Curtis JL, Sundaram B, et al.
Analysis of the lung microbiome in the “healthy” smoker and in
COPD. Plos One 2011;6:e16384.
14. Hilty M, Burke C, Pedro H, Cardenas P, Bush A, Bossley C, Davies J,
Ervine A, Poulter L, Pachter L, et al. Disordered microbial
communities in asthmatic airways. Plos One 2010;5:e8578.
15. Chambers DC, Gellatly SL, Hugenholtz P, Hansbro PM. JTD special
edition ’Hot Topics in COPD’: the microbiome in COPD. J Thorac Dis
2014;6:1525–1531.
16. Cui L, Morris A, Huang L, Beck JM, Twigg HL III, von Mutius E, Ghedin E.
The microbiome and the lung. Ann Am Thorac Soc 2014;11:S227–S232.
17. Dickson RP, Martinez FJ, Huffnagle GB. The role of the microbiome in
exacerbations of chronic lung diseases. Lancet 2014;384:691–702.
18. Rogers GB, Shaw D, Marsh RL, Carroll MP, Serisier DJ, Bruce KD.
Respiratory microbiota: addressing clinical questions, informing
clinical practice. Thorax 2015;70:74–81.
19. Frank DN, St Amand AL, Feldman RA, Boedeker EC, Harpaz N, Pace
NR. Molecular-phylogenetic characterization of microbial community
imbalances in human inflammatory bowel diseases. Proc Natl Acad
Sci USA 2007;104:13780–13785.
20. Manichanh C, Rigottier-Gois L, Bonnaud E, Gloux K, Pelletier E,
Frangeul L, Nalin R, Jarrin C, Chardon P, Marteau P, et al. Reduced
diversity of faecal microbiota in Crohn’s disease revealed by a
metagenomic approach. Gut 2006;55:205–211.
21. Agustı́ A, MacNee W, Donaldson K, Cosio M. Hypothesis: does COPD
have an autoimmune component? Thorax 2003;58:832–834.
22. Brusselle GG, Demoor T, Bracke KR, Brandsma CA, Timens W.
Lymphoid follicles in (very) severe COPD: beneficial or harmful? Eur
Respir J 2009;34:219–230.
23. Aloisi F, Pujol-Borrell R. Lymphoid neogenesis in chronic inflammatory
diseases. Nat Rev Immunol 2006;6:205–217.
24. Hogg JC, Chu F, Utokaparch S, Woods R, Elliott WM, Buzatu L,
Cherniack RM, Rogers RM, Sciurba FC, Coxson HO, et al. The
nature of small-airway obstruction in chronic obstructive pulmonary
disease. N Engl J Med 2004;350:2645–2653.
25. van der Strate BW, Postma DS, Brandsma CA, Melgert BN, Luinge MA,
Geerlings M, Hylkema MN, van den Berg A, Timens W, Kerstjens HA.
Cigarette smoke-induced emphysema: a role for the B cell? Am J
Respir Crit Care Med 2006;173:751–758.
26. Bracke KR, Verhamme FM, Seys LJ, Bantsimba-Malanda C,
Cunoosamy DM, Herbst R, Hammad H, Lambrecht BN, Joos GF,
Brusselle GG. Role of CXCL13 in cigarette smoke-induced lymphoid
follicle formation and chronic obstructive pulmonary disease. Am J
Respir Crit Care Med 2013;188:343–355.
27. Mori M, Andersson CK, Svedberg KA, Glader P, Bergqvist A,
Shikhagaie M, Löfdahl CG, Erjefält JS. Appearance of remodelled
and dendritic cell-rich alveolar-lymphoid interfaces provides
a structural basis for increased alveolar antigen uptake in
chronic obstructive pulmonary disease. Thorax 2013;68:
521–531.
28. Bonarius HP, Brandsma CA, Kerstjens HA, Koerts JA, Kerkhof M,
Nizankowska-Mogilnicka E, Roozendaal C, Postma DS, Timens W.
Antinuclear autoantibodies are more prevalent in COPD in
association with low body mass index but not with smoking history.
Thorax 2011;66:101–107.
29. Cottin V, Fabien N, Khouatra C, Moreira A, Cordier JF. Anti-elastin
autoantibodies are not present in combined pulmonary fibrosis and
emphysema. Eur Respir J 2009;33:219–221.
30. Feghali-Bostwick CA, Gadgil AS, Otterbein LE, Pilewski JM, Stoner
MW, Csizmadia E, Zhang Y, Sciurba FC, Duncan SR. Autoantibodies
in patients with chronic obstructive pulmonary disease. Am J Respir
Crit Care Med 2008;177:156–163.
31. Lee SH, Goswami S, Grudo A, Song LZ, Bandi V, Goodnight-White S,
Green L, Hacken-Bitar J, Huh J, Bakaeen F, et al. Antielastin
autoimmunity in tobacco smoking-induced emphysema. Nat Med
2007;13:567–569.
32. Rinaldi M, Lehouck A, Heulens N, Lavend’homme R, Carlier V,
Saint-Remy JM, Decramer M, Gayan-Ramirez G, Janssens W.
Antielastin B-cell and T-cell immunity in patients with
chronic obstructive pulmonary disease. Thorax 2012;67:
694–700.

986 American Journal of Respiratory and Critical Care Medicine Volume 193 Number 9 | May 1 2016
ORIGINAL ARTICLE
33. Lochner M, Ohnmacht C, Presley L, Bruhns P, Si-Tahar M, Sawa S,
Eberl G. Microbiota-induced tertiary lymphoid tissues aggravate
inflammatory disease in the absence of RORgamma t and LTi cells.
J Exp Med 2011;208:125–134.
34. Wu HJ, Ivanov II, Darce J, Hattori K, Shima T, Umesaki Y, Littman DR,
Benoist C, Mathis D. Gut-residing segmented filamentous bacteria
drive autoimmune arthritis via T helper 17 cells. Immunity 2010;32:
815–827.
35. Berer K, Mues M, Koutrolos M, Rasbi ZA, Boziki M, Johner C,
Wekerle H, Krishnamoorthy G. Commensal microbiota and myelin
autoantigen cooperate to trigger autoimmune demyelination. Nature
2011;479:538–541.
36. Lee YK, Menezes JS, Umesaki Y, Mazmanian SK. Proinflammatory
T-cell responses to gut microbiota promote experimental autoimmune
encephalomyelitis. Proc Natl Acad Sci USA 2011;108:4615–4622.
37. Sajjan U, Ganesan S, Comstock AT, Shim J, Wang Q, Nagarkar DR,
Zhao Y, Goldsmith AM, Sonstein J, Linn MJ, et al. Elastase- and
LPS-exposed mice display altered responses to rhinovirus infection.
Am J Physiol Lung Cell Mol Physiol 2009;297:L931–L944.
38. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD,
Costello EK, Fierer N, Peña AG, Goodrich JK, Gordon JI, et al. QIIME
allows analysis of high-throughput community sequencing data. Nat
Methods 2010;7:335–336.
39. Edgar RC. Search and clustering orders of magnitude faster than
BLAST. Bioinformatics 2010;26:2460–2461.
40. DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, Keller K,
Huber T, Dalevi D, Hu P, Andersen GL. Greengenes, a chimera-
checked 16S rRNA gene database and workbench compatible with
ARB. Appl Environ Microbiol 2006;72:5069–5072.
41. R Development Core Team. R: a language and environment for
statistical computing. Vienna, Austria: R Foundation for Statistical
Computing; 2008.
42. Marsland BJ, Kurrer M, Reissmann R, Harris NL, Kopf M.
Nippostrongylus brasiliensis infection leads to the development of
emphysema associated with the induction of alternatively activated
macrophages. Eur J Immunol 2008;38:479–488.
43. Park H, Shin JW, Park SG, Kim W. Microbial communities in the upper
respiratory tract of patients with asthma and chronic obstructive
pulmonary disease. Plos One 2014;9:e109710.
44. Engler K, Mühlemann K, Garzoni C, Pfahler H, Geiser T, von Garnier C.
Colonisation with Pseudomonas aeruginosa and antibiotic resistance
patterns in COPD patients. Swiss Med Wkly 2012;142:w13509.
45. Lambiase A, Del Pezzo M, Raia V, Sepe A, Ferri P, Rossano F.
Chryseobacterium respiratory tract infections in patients with cystic
fibrosis. J Infect 2007;55:518–523.
46. Zhang J, Chu S, Zhong X, Lao Q, He Z, Liang Y. Increased expression
of CD41IL-171 cells in the lung tissue of patients with stable
chronic obstructive pulmonary disease (COPD) and smokers. Int
Immunopharmacol 2013;15:58–66.
47. Podolin PL, Foley JP, Carpenter DC, Bolognese BJ, Logan GA, Long E
III, Harrison OJ, Walsh PT. T cell depletion protects against alveolar
destruction due to chronic cigarette smoke exposure in mice. Am J
Physiol Lung Cell Mol Physiol 2013;304:L312–L323.
48. Kurimoto E, Miyahara N, Kanehiro A, Waseda K, Taniguchi A, Ikeda G,
Koga H, Nishimori H, Tanimoto Y, Kataoka M, et al. IL-17A is
essential to the development of elastase-induced pulmonary
inflammation and emphysema in mice. Respir Res 2013;14:5.
49. Eppert BL, Wortham BW, Flury JL, Borchers MT. Functional
characterization of T cell populations in a mouse model of chronic
obstructive pulmonary disease. J Immunol 2013;190:1331–1340.
50. Anthony D, Seow HJ, Uddin M, Thompson M, Dousha L, Vlahos R,
Irving LB, Levy BD, Anderson GP, Bozinovski S. Serum amyloid A
promotes lung neutrophilia by increasing IL-17A levels in the mucosa
and gd T cells. Am J Respir Crit Care Med 2013;188:179–186.
51. Chen K, Pociask DA, McAleer JP, Chan YR, Alcorn JF, Kreindler JL,
Keyser MR, Shapiro SD, Houghton AM, Kolls JK, et al. IL-17RA is
required for CCL2 expression, macrophage recruitment, and
emphysema in response to cigarette smoke. Plos One 2011;6:
e20333.
Yadava, Pattaroni, Sichelstiel, et al.: Microbiota Sustains Chronic Pulmonary Inflammation
52. Di Stefano A, Caramori G, Gnemmi I, Contoli M, Vicari C, Capelli A,
Magno F, D’Anna SE, Zanini A, Brun P, et al. T helper type 17-related
cytokine expression is increased in the bronchial mucosa of stable
chronic obstructive pulmonary disease patients. Clin Exp Immunol
2009;157:316–324.
53. Taraseviciene-Stewart L, Douglas IS, Nana-Sinkam PS, Lee JD, Tuder
RM, Nicolls MR, Voelkel NF. Is alveolar destruction and emphysema
in chronic obstructive pulmonary disease an immune disease? Proc
Am Thorac Soc 2006;3:687–690.
54. Sullivan AK, Simonian PL, Falta MT, Mitchell JD, Cosgrove GP, Brown
KK, Kotzin BL, Voelkel NF, Fontenot AP. Oligoclonal CD41 T cells in
the lungs of patients with severe emphysema. Am J Respir Crit Care
Med 2005;172:590–596.
55. Marsland BJ, Yadava K, Nicod LP. The airway microbiome and
disease. Chest 2013;144:632–637.
56. Naik S, Bouladoux N, Wilhelm C, Molloy MJ, Salcedo R,
Kastenmuller W, Deming C, Quinones M, Koo L, Conlan S, et al.
Compartmentalized control of skin immunity by resident
commensals. Science 2012;337:1115–1119.
57. Molyneaux PL, Mallia P, Cox MJ, Footitt J, Willis-Owen SA, Homola D,
Trujillo-Torralbo MB, Elkin S, Kon OM, Cookson WO, et al.
Outgrowth of the bacterial airway microbiome after rhinovirus
exacerbation of chronic obstructive pulmonary disease. Am J Respir
Crit Care Med 2013;188:1224–1231.
58. Venkataraman A, Bassis CM, Beck JM, Young VB, Curtis JL,
Huffnagle GB, Schmidt TM. Application of a neutral community
model to assess structuring of the human lung microbiome. MBio
2015;6:6.
59. Bassis CM, Erb-Downward JR, Dickson RP, Freeman CM, Schmidt
TM, Young VB, Beck JM, Curtis JL, Huffnagle GB. Analysis of the
upper respiratory tract microbiotas as the source of the lung and
gastric microbiotas in healthy individuals. MBio 2015;6:e00037.
60. Shaw MH, Kamada N, Kim YG, Núñez G. Microbiota-induced IL-1b,
but not IL-6, is critical for the development of steady-state TH17 cells
in the intestine. J Exp Med 2012;209:251–258.
61. Martin B, Hirota K, Cua DJ, Stockinger B, Veldhoen M. Interleukin-17-
producing gammadelta T cells selectively expand in response to
pathogen products and environmental signals. Immunity 2009;31:
321–330.
62. Duan J, Chung H, Troy E, Kasper DL. Microbial colonization drives
expansion of IL-1 receptor 1-expressing and IL-17-producing
gamma/delta T cells. Cell Host Microbe 2010;7:140–150.
63. Trompette A, Gollwitzer ES, Yadava K, Sichelstiel AK, Sprenger N,
Ngom-Bru C, Blanchard C, Junt T, Nicod LP, Harris NL, et al. Gut
microbiota metabolism of dietary fiber influences allergic airway
disease and hematopoiesis. Nat Med 2014;20:159–166.
64. Yadava K, Marsland BJ. Lymphoid follicles in chronic lung diseases.
Thorax 2013;68:597–598.
65. Kirkham PA, Caramori G, Casolari P, Papi AA, Edwards M, Shamji B,
Triantaphyllopoulos K, Hussain F, Pinart M, Khan Y, et al. Oxidative
stress-induced antibodies to carbonyl-modified protein correlate
with severity of chronic obstructive pulmonary disease. Am J Respir
Crit Care Med 2011;184:796–802.
66. Seys LJ, Verhamme FM, Schinwald A, Hammad H, Cunoosamy DM,
Bantsimba-Malanda C, Sabirsh A, McCall E, Flavell L, Herbst R, et al.
Role of b cell-activating factor in chronic obstructive pulmonary
disease. Am J Respir Crit Care Med 2015;192:706–718.
67. Curtis JL, Freeman CM, Huffnagle GB. “B” for bad, beneficial, or both?
Lung lymphoid neogenesis in chronic obstructive pulmonary
disease. Am J Respir Crit Care Med 2015;192:648–651.
68. Brandsma CA, Kerstjens HA, van Geffen WH, Geerlings M, Postma DS,
Hylkema MN, Timens W. Differential switching to IgG and IgA in
active smoking COPD patients and healthy controls. Eur Respir J
2012;40:313–321.
69. Murphy TF, Brauer AL, Grant BJ, Sethi S. Moraxella catarrhalis in
chronic obstructive pulmonary disease: burden of disease
and immune response. Am J Respir Crit Care Med 2005;172:
195–199.
987